[
https://issues.apache.org/jira/browse/LUCENE-4759?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13575401#comment-13575401
]
Commit Tag Bot commented on LUCENE-4759:
----------------------------------------
[trunk commit] Shai Erera
http://svn.apache.org/viewvc?view=revision&revision=1444509
LUCENE-4759: remove FacetRequest.SortBy
> Remove FacetRequest.SortBy
> --------------------------
>
> Key: LUCENE-4759
> URL: https://issues.apache.org/jira/browse/LUCENE-4759
> Project: Lucene - Core
> Issue Type: Improvement
> Components: modules/facet
> Reporter: Shai Erera
> Attachments: LUCENE-4759.patch
>
>
> FacetRequest.SortBy lets you specify two sort-by: ORDINAL and VALUE. While
> VALUE is the default (and breaks ties by ordinal), it's not very clear what's
> the use of SortBy.ORDINAL.
> In practice, if you choose to do that, you'd get the first K categories that
> are the descendant of the requested one, from smallest to highest, or vice
> versa. But that seems quite useless ... someone could just traverse the
> counts array (for instance) and filter out all counts==0? Or also, someone
> can write a FacetResultsHandler which does that...
> My motivation to remove that is to reduce the number of PQ combinations we
> have: MinValue, MaxValue (SortBy.VALUE, SortOrder.ASCENDING/DESCENDING) and
> MinOrdinal, MaxOrdinal. Now there are 4 PQs and I'd like to separately split
> them out to PQs that handle int vs float. Because today these PQs call
> Double.compare(), which you need to for floating-point values, but is just a
> waste for integer values.
> So removing SortBy will both simplify the API and halve the number of PQs we
> need to write. Plus ... it doesn't seem such a useful option, to let the user
> even spend 10 seconds to read the differences between VALUE and ORDINAL.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
